Adaptive Diurnal Prediction of Ambient Dry-Bulb Temperature and Solar Radiation

Abstract

This paper presents a new adaptive weather-prediction model that can be used for on-line control of HVAC and thermal storage systems. The model can predict external dry-bulb temperature and solar radiation over the next 24 h. Because a building with a fabric thermal storage system has a slow response to thermal loads, a predictive controller is essential to operate the building and associated plant installation to respond effectively to external climatic conditions ahead of time. Three prediction methods are investigated in the paper: a pure stochastic method, a combined deterministic-stochastic method, and an expanded method for short-term temperature forecast. It has been found that the combined deterministic-stochastic method is simpler and gives the smallest prediction errors. For the prediction of solar radiation, a deterministic model is proposed. The proposed prediction algorithms for temperature and radiation are simple and efficient to conduct on a supervisory PC to predict hourly temperature and radiation profiles over the next 24 h. Updating temperature forecasts using observations available with time is also investigated in this paper.